We are trying to use file-polling in one of our requirements.But had one doubt.
What if all the files will come with the same name.Will it override the existing file even the file is data is not processed or Any other issue might happen due to same file name.
How to proceed with the processing with the data in each of the files in order to be sure there is no data loss.
If you are looking for a best practice to avoid this kind of file overwrite situations in the file polling pickup process you should consider the design of clients/source system request them send the file names with datetime stamp (as a mandate). Also you can consider configuring actual vs working folder to avoid the files overwrite and processing.
Please make a note using datetime inclusion in the file names and it should avoid the data loss worries.
It depends on the requirements.
As rmg mentioned the best way is to make sure that the source system generates files with different names.
Do you know with what frequency will you receive the files?
Another (not fail-proof) option would be to setup the File Polling Interval to be less than the mentioned frequency.
This way the files will not have the time to overwrite each other.